Mel sub-band filtering and compression for robust speech recognition.
Babak NasersharifAhmad AkbariMohammad Mehdi HomayounpourPublished in: INTERSPEECH (2007)
Keyphrases
- speech recognition
- subband
- image compression
- cepstral coefficients
- noisy environments
- hidden markov models
- wavelet transform
- high frequency
- pattern recognition
- filter bank
- bit rate
- high pass
- language model
- speech synthesis
- linear prediction
- wavelet coefficients
- feature vectors
- speech signal
- automatic speech recognition
- discrete wavelet transform
- wavelet decomposition
- low pass
- multiresolution
- locally adaptive
- low frequency
- compression algorithm
- speech recognizer
- speaker independent
- frequency domain
- neural network
- speech recognition systems
- lossy compression
- video signals
- speaker identification
- speech recognition technology
- speaker adaptation
- wavelet subbands
- frequency band
- wavelet domain
- compression ratio
- signal processing
- principal component analysis
- image processing